Continued analysis of data on a nationally representative sample of 2,143 families is proposed. Work to date has provided the most accurate (and in some cases the only) incidence rate statistics on physical violence between family members, including, child-abuse, wife-beating, husband-beating, violence by children toward each other, and violence by children towards parents. Further analysis is needed because the work to date has been largely descriptive analyses of incidence data and analysis of two-variable relationships, and because a number of important issues are still to be investigated. Eleven needed analyses are described. Examples of the questions which these are designed to answer are: (l) What is the relationship between the balance of power between husband and wife and the use of physical violence? How is this affected by such factors as the social class of the couple, and the degree to which the wife is economically independent of the husband? (2) What accounts for group differences in rates of family violence? Among the group differences to be studied (because work to date shows important differences) are the region of the country, socioeconomic status, and race. (3) Is involvement in a primary group community of friends and kin associated with a lower level of family violence, and why? (4) We have found a strong relationship between growing up in a violent family and being violent to one's spouse and children. What about those who do not follow this pattern? Do they have family patterns which will suggest ways to break the generation to generation cycle of violence? The importance of finding answers to questions such as these stems from the extremely high violence rates demonstrated by this project. This adversely affects the quality of family life and is part of the explanation to the high level of violence in the society. There is clearly a need for continued research which can suggest policies to reduce family violence.